—Freakonomics79
In the above quotation, Levitt and Dubner argue that corporate fraud is rampant, but they say it’s hardly ever detected. They might forgive a reader for wondering how they can know that these misdeeds are so prevalent if no one ever finds out about them. This claim is really just an assertion of what they think might be happening; no evidence is offered because—conveniently—no evidence is possible. It’s a perfectly unfalsifiable claim.
Popular belief in rampant corporate fraud has real consequences in that it creates pressure for changes in government policy. Reflecting popular opinion that penalties against corporate fraud were too low, in 1991 the U.S. Sentencing Commission—the federal agency responsible for setting the penalty guidelines used by judges—raised median corporate fraud penalties by over twenty-fold. I fought vehemently against this move while I was chief economist at the Sentencing Commission during the late 1980s, but only succeeded in getting it temporarily delayed.
At the time, the penalties seemed small compared to the losses imposed on customers due to corporate fraud. In the late 1980s, half the corporate fraud convictions resulted in sanctions of less than fifteen cents per dollar lost from the fraud, with the average fine amounting to only seventy-five cents per lost dollar. This was much lower than the relative penalties meted out for other corporate crimes such as environmental pollution, where half the fines were $3.71 or more per lost dollar.80
Yet, the research I did with Jonathan Karpoff, a professor at the University of Washington, convinced me that there is a solid economic reason for setting corporate fraud penalties lower than for other misdeeds such as environmental crimes. When convicted of fraud, companies face collateral penalties related to their loss of reputation. And these reputational penalties usually do not apply to other kinds of corporate crime. When firms defraud their customers or don’t deliver what they promise, customers will stop buying their products or will insist on lower prices that factor in the extra risk. In contrast, aside from a small minority of activists, most customers will not reject a company’s products because it was convicted of an environmental or similar crime that does not directly affect the customers’ purchases.81
The reputational penalties suffered by a firm accused of fraud are substantial, even before the company is convicted. The market value of firms accused of fraud fell by $61 million during 1980s.82 On average, only 6.5 percent of this decline reflects legal costs, and just 1.4 percent is accounted for by penalties and fines. The rest of the drop reflects expectations of reduced sales and earnings.83 This contrasts sharply with the drop in stock prices when a firm is accused of an environmental crime, in which case virtually the entire decline reflects the firm’s legal costs and penalties. In essence, reputational penalties, on average, are non-existent for environmental crime. People may not like hearing that a company pollutes, but overall they simply will not stop buying the firm’s products based on environmental misdeeds.
Thus it is primarily the fear of legal penalties, not reputational ones, that deters environmental crimes. But this is not the case for corporate fraud. As previously noted, at the time when corporate fraud penalties were increased twenty-fold, government penalties for fraud averaged just seventy-five cents per lost dollar. When this is combined with the average reputational penalties in the form of declining sales, earnings, and stock prices, we find that the average total penalty due to a fraud accusation was already 11.5 times greater than the loss imposed on customers.
This is not mere academic curiosity; raising the criminal penalties for fraud too high can hurt customers and damage the economy. Large fraud penalties force companies to spend more money on guaranteeing product quality. While everyone values greater assurance that they are getting higher quality products, not everyone wants to pay more for this guarantee. Just as different people take different levels of risk in their everyday lives, different people are willing to pay different amounts for insuring quality. If a person wants the maximum guarantee that a car will perform as it should, he can shell out lots of money to buy a new Lexus. But for some people, price is more important than more quality assurance; they are willing to face a high risk of a car breaking down in return for paying a low price for it. As firms increase quality guarantees to avoid high fraud penalties, prices rise as well. Suppose that all firms were forced to make cars as dependable as a Lexus. We’d end up with very few accusations of fraud, but we’d also have many people who could no longer afford a car.
Aside from fraud cases, reputations factor unexpectedly into other aspects of business management. Let’s look at just one everyday example—the franchising of gas stations. There are a lot of advantages to franchising. Franchises tend to be well-run because the owners, with a direct interest in profitability, are right there on the premises overseeing operations, as opposed to sitting in some remote corporate headquarters. Having well-managed franchises also burnishes the reputation of the franchisor corporation, whose logo, of course, is attached to the franchise.
So why aren’t all gas stations franchised? The answer becomes apparent if we look at areas where corporations prefer not to franchise. Corporate headquarters tend to operate directly the company’s own gas stations in areas without many repeat customers, for example, next to a superhighway. The problem with a franchise in this situation is that the owner has less of an incentive than usual to operate the business because satisfied customers are unlikely to return to the station anyway. If customers get cheated or are subject to poor service, however, they may stop frequenting that chain in many other places, thus damaging other franchises carrying the same logo. The franchisee that actually provided the poor service will not suffer much of the consequences of its own poor service.84
Reputations play an important, though often unacknowledged, role in society. Notwithstanding the high-profile corporate and political scandals that will always occur from time to time, concern over reputations helps keep people honest; the loss of a good reputation can easily devastate the future prospects for a politician, corporation, or an individual, as we have seen. As long as this remains true, reputations will continue to function as a vital component of our free market democracy.
3
Government as Nirvana?
People frequently call for government intervention in the economy whenever the market is believed to be acting imperfectly.1 Implicitly, the comparison is between the flawed way the market actually works on the one hand, and a nirvana-like state of government-run perfection on the other. Do distortions ever develop in the free market? Of course they do. Few people would argue that the market is flawless. But it’s a long leap from showing that such imperfections exist to proving that they would be solved or even mitigated by government intervention. In fact, government intrusion in the economy tends to result in more inefficiency, unfairness, and even predation than we would find in a completely free market.
C’mon and Take a Free Ride
One of the most common methods of government intervention is to encourage certain beneficial actions through subsidies. Recall our earlier discussion about LoJack, the car antitheft device. LoJack’s social benefit—the overall drop in car thefts we would allegedly see due to the fear among thieves that any car might have a LoJack—was said to be ten times larger than the benefit for any individual LoJack users.2 But the government would have to subsidize the device in order to get car manufacturers to install it in enough vehicles to capture the full social benefit.3 Advocates of such a subsidy used this kind of argument to justify forcing insurance companies to give discounts to LoJack users.4
Remember the counter-argument, however. Porsche could solve this problem for its customers by putting LoJacks on its own vehicles, and thieves would consequently steal fewer Porsches and more of every other kind of car. But this would harm everyone who doesn’t drive a Porsche. In that case, instead of being rewarded with a subsidy for installing LoJack, Porsche should instead have to pay a tax on it.
Here we see that having
a subsidy when there really should be a tax—or visa-versa—is worse than doing nothing. If we subsidize a harmful activity, we end up encouraging it. If we tax a beneficial activity, we inadvertently discourage it. And if a subsidy or tax is too large or too small, we create market distortions. Smoking may be unhealthy, but excessively high cigarette taxes will create a black market in cigarettes. Similarly, growing wheat may be vital for feeding the country, but over-subsidizing it will induce farmers to stop growing other crops. In the case of LoJack, if the device has no significant impact on auto theft rates, as attested to by car manufacturers, insurance companies, and recent academic research, then the product should not have any special subsidies or taxes at all.
Even if the benefits of LoJack were confirmed and the proper subsidy could be precisely calculated, there is another problem: trusting the government to approve the right subsidy. Politicians representing districts where LoJacks are made would probably support a subsidy far in excess of the correct amount, while politicians representing competitors’ districts would likely seek to reduce the funding.5
A similar example involves concealed handguns. Those who carry permitted concealed handguns deter criminals from attacking others in their localities because criminals fear that any potential victim may be carrying a gun. Despite this social benefit, it’s not clear that the government should subsidize concealed handguns. As with most cases of government subsidies, the problem lies in the difficulty of calculating how much the subsidy should be. Even though a clear social benefit exists, figuring out the subsidy would require complex estimates of the number of crimes that would have taken place if concealed weapons weren’t allowed, as well as calculations of the monetary and social cost of these crimes that never occurred. Keep in mind that the politicians and civil servants drafting these regulations are rarely experts in the areas they oversee. As for concealed handguns, all states in fact discourage permits by charging a mandatory licensing fee that is usually far above the administrative cost of issuing the permits. Far from being subsidized, concealed weapons permits are, in effect, taxed.
There is another similarity between LoJacks and right-to-carry laws. Some states allow residents to carry a gun openly. This certainly benefits the person carrying the gun—not many criminals would attack someone they knew to be armed. But open-carry doesn’t produce the same social benefit as concealed handguns do. Although open-carry discourages crime around the person openly carrying a gun, criminals may simply go after other people without guns, just like installing LoJacks in Porsches would encourage criminals to steal other cars. So open-carry has a private benefit for gun carriers, but it also may increase the rate of attacks on people without weapons.6
The late economist Milton Friedman noted a further complication in subsidizing products or practices that create social benefits: subsidies may not even be necessary because people often have enough individual incentive to do the desired action without any government intervention. Take Friedman’s example—education. Let’s assume by the time a student enters high school, his education has instilled in him positive values such as support for democracy and the avoidance of criminality. Perhaps this indicates that education should be subsidized, since more overall schooling benefits society by strengthening the general commitment to democracy and reducing crime. Yet, education also offers the private, individual benefit of increasing a student’s future job prospects. As Friedman pointed out, we can get the social benefit of education without subsidizing it because students will attend school to obtain this personal benefit.7
The debates over government subsidies are closely linked to the “free-rider problem” in economics. Suppose a person spends time or money on something that incidentally benefits others, but these beneficiaries pay little, if any, of the costs. Such free-riding situations tend to result in a dearth of the beneficial action, since many people will hope to get the benefit for free through others’ actions. This situation often elicits demands for state subsidies to encourage the valuable activity. This applies to the use of crime prevention devices like LoJack as well as actions like charitable giving and volunteerism, which are sometimes effectively subsidized through tax deductions.
There are two questions involved in free-riding problems: Do others benefit from your action? And if so, is it possible to prevent them from getting this benefit? When it’s impossible to exclude others from the benefit, a free-riding problem is likely to develop. For example, if LoJack were actually effective, a free-riding problem could easily appear, since those without LoJack would hope thieves will stay away from their own cars in fear that it might have the device. Enough free-riders might then stop buying LoJack that the social benefit—the overall drop in car thefts—would disappear.8
Another example of free-riding was seen during the early development of radio. Today, virtually everyone takes it for granted that advertising is a sensible way to finance radio broadcasting. Few people realize that free-riding problems initially seemed almost insurmountable in providing radio service. Because no one could figure out how to make listeners pay, radio hosts and entertainers usually had to work for free.9 For over twenty years, broadcasting primarily involved hobbyists and a few public service transmissions by government stations.
Some people doubted there was any way to make listeners pay. In 1922, Herbert Hoover, then Secretary of Commerce, declared: “Nor do I believe there is any practical method of payment from the listeners.” 10 Others assumed that radio transmissions would eventually be funded by paying subscribers, but no one could devise a method for limiting broadcasts to subscribers’ receivers. Consequently, some believed the government would have to provide the service. In 1922, Popular Radio magazine claimed that radio was “essentially a public utility” and discussed using city telephone wires to sell broadcasts to subscribers—in other words, providing radio service over the phone.11
So what happened? Did private businessmen throw up their hands and invite the government to run the industry? Was society denied the benefit of radio because no one could solve the free-riding problem? Of course not. The problem was eventually resolved in 1922 when AT&T discovered it could make money by selling radio advertising airtime. In hindsight, it’s hard to believe that private radio almost died in its infancy because people couldn’t figure out how it could make money. And it’s a good thing that the government decided not to turn radio into a subsidized enterprise, since it is highly unlikely that the state would have distributed payments as efficiently as advertisers do.
With enough at stake, companies find amazingly creative ways to solve free-riding problems. Government subsidies only deaden the incentive to discover these solutions. For example, many analysts used to regard beekeeping and apple farming as a classic free-riding problem.12 Apple blossom nectar provides food for bees, which pollinate the blossoms as they gather the nectar. Economists feared that apple growers were free-riding from neighboring beekeepers, and that this would eventually result in too few apples as well as too little honey. But apple growers and beekeepers devised a number of solutions to this problem. The most obvious method was for apple farmers to set up their own bee-hives. 13 However, as Steven Cheung showed, the more common solution was for apple growers and beekeepers to create markets where they transacted regularly. The Yellow Pages in rural Washington State developed long listings of pollination services available to farmers. Yet, despite the evolution of free market solutions, beekeepers cited the existence of this potential free-riding situation as an excuse to lobby the government to implement a honey price support program. The program began modestly in the early 1980s, but quickly grew into a massive subsidy scheme costing U.S. taxpayers around $100 million annually by 1984.14
Government attempts to solve free-riding problems are typical examples of the state’s economic inefficiency. In fact, we’ve grown so accustomed to the inadequacies of government that we typically use different standards for evaluating private and public spending. For example, economists deem private markets
to be efficient when the cost of an additional unit of some product reflects how much buyers value it. But when is that ever a consideration in government spending? Just look at tax payments generally: the top 5 percent of income earners pay 57 percent of federal income taxes, while the bottom 50 percent pay just 3.3 percent.15 In a democracy, those who provide little of the government’s income have more of a say—in the form of their combined votes—over how to spend government funds than those who provide most of the money. Of course, this is not to suggest that the votes of big taxpayers should be weighted more than those of smaller ones. It’s just to point out that government spending is inherently inefficient because those who actually pay for most government services are not the ones who determine how the money is spent.
This helps to explain why government intervention is so often inefficient. 16 In private markets, you can’t get people to pay more for a product than they value it. If the asking price is too high, they simply say “no.” But there is no similar limitation on the government, which pays for things by levying taxes. And taxes are coercive—you can’t refuse to pay taxes just because the government is paying more for something than you value it.
Take government programs such as flood insurance. These aren’t like traditional private insurance programs where people are charged according to the risk they represent. Private insurance companies closely match the premiums to the risk level or else they quickly go out of business. If they charge too much, their customers go elsewhere. If they charge too little, they lose money. While things have improved somewhat in recent years, for decades the government insurance programs charged everyone the same amount regardless of risk. The government charged the same flood insurance premiums for beachfront houses as it did for homes in the middle of the desert. As late as June 30, 2005, the Congressional Research Service was still reporting on the “repetitive loss problem,” where people in high risk flood areas paid such low flood insurance premiums that they would keep on rebuilding only to have their homes repeatedly swept away.17 As USA Today noted, “One Houston property valued at $114,480 has filed for losses 16 times and received $807,000 in total payments.”18
Freedomnomics: Why the Free Market Works and Other Half-Baked Theories Don't Page 9
